Corrigendum to NZJFS22(1) 39-53: How will New Zealand's forests respond to climate change? Potential changes in response to increasing temperature.

The paper which this Corrigendum refers to is available here: How will New Zealand's forests respond to climate change? Potential changes in response to increasing temperature

Three, mature, natural forest ecosystems on typical sites north and east of Lake Superior, Ontario, Canada, were contrasted in terms of their content and distribution of organic carbon (C). Total carbon reserves were lowest in an approximately 62-year-old jack pine (Pinus banksiaria Lamb.) stand on a sandy outwash soil, highest in an old-growth (up to 300 years) sugar maple (Acer saccharum Marsh.) stand on an upland till soil, and intermediate in an approximately 110-year-old black spruce (Picea mariana [Mill.] B.S.P.) stand on a shallow upland till soil.

Annual net carbon fixation was substantially higher in the sugar maple than in the jack pine stand. Allocation of assimilated carbon varied: in the sugar maple stand, greater amounts went to the formation of leaves and fine roots; in the jack pine stand, new carbon was apportioned in the order-woody biomass > leaves > roots > flowers and fruits. Calculations of residence time (taking into account litter, fine root, and solution inputs) suggest that turnover of carbon is approximately three times more rapid in the sugar maple than in the jack pine forest floor. Residual carbon (taking into account inputs, accretion, and losses) for return to the atmosphere was likewise approximately three times greater in the sugar maple than in the jack pine forest floor. If not for respiration, all subsoil organic carbon could be accounted for by leaching inputs less outputs in approximately 100 years, indicating rapid turnover of carbon in the subsoil as well.

Removal of carbon by different harvesting methods was calculated. In general, conventional (shortwood or tree-length) systems would result in removals of approximately 20% of the total (organic) carbon reserve from either the maple or the spruce site, with more intensive harvesting systems removing up to 32-35%. In contrast, conventional harvesting in the pine stand would remove approximately 33% of the total carbon, with more intensive regimes removing around 38-44%. The impact of site preparation method varied. Jack pine and black spruce sites could be susceptible to fertility loss as a result of full-tree harvesting because of the large amounts of carbon and nutrients stored in the forest floor rather than in the mineral soil.

Data from a series of experiments illustrated the importance of organic carbon in influencing a range of key determinants of plantation productivity on podsolised sands, which lack a significant inorganic colloidal phase. Organic carbon levels affected soil nitrogen reserves, nitrogen dynamics, phosphorus availability, and cation exchange capacity.

Evidence from two experiments indicated that the organic carbon in podsolised sands is highly dynamic and sensitive to management operations which influence organic carbon inputs, decomposition rates, or both. Weeds can help maintain organic carbon reserves after clearfelling, particularly where logging residues have been burnt. Retention of above-ground logging residues also helps to maintain organic carbon reserves. Most harvesting and site preparation operations result in loss of a labile carbon pool (representing approximately 30% of total carbon). This pool can be buffered by residue retention and weeds in the period before significant litter inputs from the new crop, but in any case will be replenished once these inputs are resumed. As such, the impact of management on this pool is likely to be transient. Long-term reductions in soil carbon and associated properties are likely only where management operations result in loss from the recalcitrant carbon pool (representing approximately 70% of total carbon). Since such material decomposes extremely slowly, only site preparation operations such as surface soil scalping or the use of high-intensity fire are likely to result in significant short-term losses of this fraction.

Using a model of a spruce forest ecosystem, comparisons of soil carbon were made after conventional (stem only) harvesting, harvesting of stems and branches, whole-tree (above ground) harvesting, and no biomass removal at clearfelling. Parameter values in the simulations corresponded to conditions and regulations in Swedish forestry.

As could be expected, soil carbon was less after intensive harvesting than after conventional harvesting, but the differences were rather small. With no removal of biomass, soil carbon increased substantially over the 300 simulation-years, but it decreased over time under all harvesting alternatives. Productivity (measured as harvested stem biomass) decreased with increasing harvesting intensity, whereas total harvested biomass increased with increased harvesting intensity in more productive stands. The results were similar to earlier model studies, and indicate that the major effects on soil carbon in forests come from conventional harvesting, and that increased utilisation of branches and needles for energy purposes is likely to have a relatively small additional impact.

Most simplifications of the model are such that the effects of intensified harvest on soil carbon in the field are likely to be smaller than the model suggests. We conclude that the fears that whole-tree harvesting will lead to substantial decreases in the amount of soil organic matter probably have been exaggerated. Given that the additional harvested biomass is used to replace fossil fuels, intensified harvesting for energy purposes in managed forests probably has only small effects on atmospheric carbon dioxide levels.

Three studies explored the potential of the conventional radiocarbon dating procedure combined with the analysis of the degree of incorporation of ¹⁴C derived from nuclear

weapons testing (¹⁴C-bomb) during the early 1960s, for determining carbon turnover in forest soils.

The first study examined the gross annual turnover rate of soil carbon within the tree rooting zone of a mixed deciduous oak/ash/birch woodland (Meathop Wood) situated on an acid brown earth soil overlying carboniferous limestone. The turnover was estimated by dividing the soil carbon content of several soil layers (O + O_f 0-5,5-10,10-15,15-25,25-35, and 35-50 cm depths) by their respective mean carbon age (carbon mean residence time) derived from the ¹⁴C analyses. The total turnover of carbon, estimated as 3860 kg C/ha/yr, very closely agreed with the estimated total carbon inputs to soil as litter components (leaf litter, twigs, flowers and fruits, herb layer throughfall, and stem flow plus root decay) of 3895 kg C/ha/yr. The latter data were derived from intensive site studies carried out during the International Biological Programme. Close agreement between the estimate for carbon turnover derived from the ¹⁴C data and the estimate for the annual total litter input is considered to validate the use of the isotope approach. It is also clear that different components within the litter layer and roots within the mineral soil layers, have different ¹⁴C-bomb signatures, thus indicating the potential to determine mean residence times of different forest litter materials within the O-horizon or roots within the mineral soil.

The second case study investigated the effects of birch on carbon dynamics in acidic heather moorland soils. Birch has gained the reputation of being a soil improver and has been shown to increase earthworm numbers, pH and extractable calcium, and mineralisable nitrogen, and significantly decrease the C:N, C:P, and C:K ratios in surface soils. Though these changes may be induced by a number of differing and interacting processes, an increase in the rate of soil organic matter decomposition, through enhanced biological activity, was thought to be a dominant factor. Using a chronosequence of sites from heather moor to 90-year-old birch stands at a single site in Scotland showing characteristic changes in soil properties, ¹⁴C measurements were made in 1976 on different soil horizons down to 40 cm depth. These showed a clear pattern of increasing ¹⁴C enrichment and the isotope penetration deeper into the soil profile with increasing stand age. These results accord with the hypothesis that there is an increasingly rapid turnover of the moorland soil humus and its partial replacement with younger birch-derived organic matter, with increasing birch stand age up to approximately 40 years.

The potential use of these ¹⁴C techniques in research on the dynamics of carbon in upland and high latitude soils containing high amounts of organic matter, in respect to the possible effects of global warming, has been suggested by studies of brown earth soils at four sites in an altitudinal gradient on the Pennines in northern England. Results showed that the soil carbon at the coolest site at 747 m had lower ¹⁴C enrichment values than that at the warmest site at 425 m. The sites differed in mean annual temperature by approx. 2°C. Furthermore ¹⁴C-bomb, indicating younger contemporary carbon, had penetrated only to c.3 cm at the highest site but had been incorporated to >5 cm at the lowest site. The combined results indicated that soil carbon turnover is much slower at the highest than at the lowest altitude site.

These ¹⁴C techniques might be used to partially validate computer models of carbon dynamics in forest ecosystems and at different scales of resolution (process, ecosystem, and landscape) in environmental studies.

General areas of research needed for soil carbon include, but are not limited to, (1) effects of forest management, (2) effects of climate change, and (3) effects of elevated carbon dioxide. The research requirements to investigate these topics include: (1) a coordinated study with specific protocols for sampling and analyses to either confirm or negate the results of a recent review which showed little average change in soil carbon with harvesting, but increased soil carbon with improved nutrient status; (2) processlevel studies involving temperature and moisture manipulations combined with gradient studies; (3) studies of litter quality effects (including roots) using actual incubation and/or litterbag experiments to assess potential litter and soil carbon changes with increasing carbon dioxide; (4) a realistic soil carbon fractionation method to complement all new studies; and (5) studies of factors affecting soil carbon dioxide partial pressures (pC0₂) along with an evaluation of the effects of soil pC0₂ on tree seedling root growth.

The problem of allocation of a plant's carbon resources is basically an evolutionary one in which long-term reproductive success is the goal. Analyses of which features maximise seed production have been done on the basis of root: shoot: seed allocation and height growth: seed production. More complex models which focus on many-species interactions along a resource continuum have also been tried. More commonly, however, the problem of allocation is defined in terms of functional (physiological) properties of the plant, with nitrogen availability as a dominant factor. Within this category, some models rely directly on empirical information, others derive root:shoot allocation from various physiological principles such as transport resistance, balance between carbon and nitrogen uptake, or balance between carbon assimilation and consumption. In some cases allocation is calculated from an optimisation scheme. In view of the criteria that should be satisfied by allocation models, we conclude that today there are no allocation models that satisfy all requirements.

A special section containing a collection of 7 papers presented at this workshop. The papers were related to 3 themes: the effects of forest management on plant-level carbon allocation in forest ecosystems; the effects of forest management processes that regulate carbon cycling in the forest floor and mineral soil; and the potential for, and constraint on, sequestering carbon in forest ecosystems.

Cunninghamia lanceolata (Lamb.) Hook. (Chinese fir) is considered one of the most important trees in central-southern China. In China it has been cultivated as a timber species for over 1000 years; however, the species does not appear to have been planted much outside China and Taiwan. Physical, mechanical, and drying properties of three stands of New Zealand-grown C. lanceolata were assessed and anatomical and pulping studies were reviewed.

Results clearly showed that mechanical and physical wood properties of New Zealand-grown C. lanceolata are numerically lower than those of native-grown (Chinese/Taiwanese) C. lanceolata. The main factor is a lower basic density resulting in reduced strength; shrinkage, however, appears to be fairly constant. In comparison to Pinus radiata D.Don, C. lanceolata wood is of lower density and therefore not as strong. Shrinkage is similar for both species and there is little degrade under conventional kiln schedule or air drying. Drying rates are similar to P. radiata.

It appears that owing to its low basic density C. lanceolata would be unsuitable for heavy structural uses. However, its dimensional stability, ease of drying, and reputed durability would allow it to be used in applications such as weatherboarding, panelling, and joinery.

A Eucalyptus delegatensis R.T.Baker provenance trial was evaluated at age 8 years for growth rate, internal checking within growth rings, and other wood properties. Internal checking, which occurs on drying, severely restricts the use of E. delegatensis for solid-wood products. Relationships were investigated between internal checking variables and the other wood properties (basic density, heartwood content, and moisture content) and stem diameter.

Neither the size nor the frequency of internal checks showed any marked association with any of the other disc variables or combination thereof. Internal checking features, however, differed strongly between two regional provenance groups, the Tasmanian provenances having less than half as many as the mainland Australian provenances. Heartwood was markedly less and basic density averaged 9kg/m³ higher in the Tasmanian material. The Tasmanian provenances were also superior, albeit marginally, in diameter growth. Accordingly, Tasmanian provenances are recommended over mainland ones for New Zealand plantations, even though tree form is not as good.

Eucalyptus delegatensis R.T.Baker provenance trials, at two sites in New Zealand, were assessed at age 8 years for growth (diameter) and form (primarily stem straightness). Tasmanian provenances overall had slightly larger (p < 0.1) diameters than Australian mainland provenances, but were significantly (p < 0.0001) poorer in form than mainland ones. The results for diameter growth at age 8 were in contrast to those for earlier (age-3) height. In both results at age 3 on the two New Zealand sites and the published results of the same material on four trial sites in south-eastern Australia, Victorian provenances had clearly excelled Tasmanian provenances. Hence, Victorian provenances that show rapid early growth may later be overtaken by Tasmanian provenances.

New Zealand seedlots (commercial lots and open-pollinated families) showed, on average, modest diameter but good form, in line with their predominantly New South Wales origins. The families varied strongly in both diameter and form.

The performances of Mc Arthur's models, that use screen temperature (T) and relative humidity (H) to predict the moisture contents of dead fine fuel (FFM) in eucalypt forest and grassland, were assessed when they were applied to (i) six common types of dead eucalypt and pine fine fuel exposed to atmospheric conditions in a meteorological screen and (ii) dead needle fuels in canopies and litter of Pinus radiata D.Don plantations.

In the screen, diurnal range of FFM in pine needles was wider and reached lower afternoon values than in other fuels. When H was within the domains applicable to model inputs, the moisture contents of both pine and eucalypt fuels were, to varying extents, under-predicted by McArthur's models. The predictions of the FFM model developed for control burning operations in eucalypt forest (the CBEF model) were most closely correlated with observed FFMs

Site index is used as a measure of productivity for large plantation forests. Although site index had been calculated in less than half of the compartments in Kaingaroa Forest, data were fairly evenly spread. A Geographic Information System (GIS) was used to produce a contour map of site indices associated with compartment centre points. The limitations of estimation techniques within the GIS were highlighted by the difficulty of predicting values between contour lines. Instead, geostatistics, a statistical interpolation method, was adopted as it can estimate local values from data that varies spatially.

The variogram for site index in Kaingaroa Forest was fitted by a linear model up to 25 km. The parameters of this model were used in estimation (kriging) procedure. Values for 757 compartments were predicted, ranging from 18.8 to 34.3 m. The standard error ranged from 1.6 to 3.6 m, with a mean of 1.7 m. A jack-knifing procedure showed estimates to agree well with actual values.

It was concluded that linking a GIS with geostatisties allowed more effective use to be made of the GIS.

The effects of six vegetation types on landslide densities on Tertiary bedrock terrain were examined before and after Cyclone Bola struck the East Coast region of the North Island of New Zealand in March 1988.

Indigenous forest and exotic pine plantations more than 8 years old provided the best protection against the formation of shallow landslides, both before and during Cyclone Bola. Regenerating scrub and exotic pines 6-8 years old provided an intermediate level of protection. Greatest damage occurred on pasture and in areas of young (<6 years old) exotic plantations where canopy cover was negligible and root development limited.

Sites under older vegetation types with a closed canopy (indigenous forest and plantations of exotic pines >8 years old) were four times less susceptible to landsliding during Cyclone Bola than those under regenerating scrub and exotic pines 6-8 years old, and 16 times less susceptible than those under pasture and young exotic pines (<6 years old).

Blends of tall oil pitch (TOP) and petroleum bitumen were studied as potential binders for road surfacings. The physical properties (viscosity, penetration, softening point) of TOP/bitumen blends (up to 30% w/w TOP) were similar to those of petroleum bitumen alone. Resistance to oxidative hardening likely to occur in asphalt concrete manufacture was examined using the standard rolling thin film oven test. The blended binders showed a drop in 25°C penetration value equivalent to that found with petroleum bitumen of the same initial penetration. No separation of the TOP/bitumen blends was observed even after 3 days' storage at 135°C. Thermogravimetric analyses in air showed the onset temperature for weight loss to be lower (~170°C) than that for bitumen (~210°C). However, the flash point of blends up to 30% w/w TOP remained well above the lower limit allowed for petroleum bitumens.

New Zealand material in Japan continues to be sawn predominantly for packaging, in contrast to other supplies which are sawn overwhelmingly for construction uses. Overall recovery percentages for New Zealand material sawn (for packaging) in Japan are of the order of 64-68% which is much higher than the national average in New Zealand of approximately 50%. The higher recovery in Japan can be explained by: the JAS under-estimate of log volume; the volume of large squares in the raw material input; the high raw-material cost which has instigated slow, accurate, finer kerf sawing; the combination of products and dimensions produced.

The Japanese sawmilling industry is adaptable, adjusting to rising raw material costs and reduced supplies by adopting strategies such as diversification into value-added processing, the establishment of industrial complexes, downstream integration such as housing construction, and real estate.

Annual carbon uptake by the 1.24 million ha of plantation forest in New Zealand was calculated from detailed information provided to the Government by private owners on the age and volume of the timber resource, a national database of wood density variations, models of the allocation of biomass to tree and forest components other than stems, and estimates of roundwood removals derived from annual Government surveys of sawmills, chip mills, and other wood product mills, as well as export data.

The plantation forests of New Zealand stored approximately 4.5±0.8 million tonnes C in the year between 1 April 1988 and 1 April 1989, increasing total plantation carbon storage to approximately 88 million tonnes C in April 1989. Without harvest, the average annual carbon uptake of the New Zealand plantation estate between 1988 and 1989 would have been approximately 6.4 tonnes C/ha. Plantation roundwood removals were equivalent to 2.7 tonnes C/ha, so that average carbon storage was approximately 3.6 tonnes C/ha. Some harvested carbon is stored in wood products, and additional carbon may be stored in the mineral soil, but these quantities were not included in our estimates. The annual storage of carbon in the New Zealand plantation estate in 1988-89 was equivalent to approximately 70% of total New Zealand fossil fuel emissions, but was <0.1% of total global fossil fuel emissions.

The high annual rate of carbon uptake by the New Zealand plantation estate is a consequence of a large area of new plantings initiated in the 1970s and 1980s. Without continued new plantings, the net annual rate of carbon uptake by New Zealand plantation forests will rapidly approach zero.

A trial designed to quantify the reduction in Pinus radiata D.Don seedling growth caused by competition from a range of important weed species was established at Rotorua, a moist North Island site, and at Rangiora, a South Island site with low summer rainfall. At both sites, P. radiata seedlings were grown on their own and with either herbaceous broadleaves (a volunteer mixture of species from which grasses were excluded), Cytisus scoparius L. (broom), or Ulex europaeus L. (gorse). Trees were also grown with Buddleja davidii Franchet (buddleia), Holcus lanatus L. (Yorkshire fog) plus Lolium multiflorum L. (Italian ryegrass), and Cortaderia selloana (Schult) Asch. et Graeb. (pampas) at Rotorua and with Agrostis capillaris L. (browntop) at Rangiora. Resource (nutrient and water) levels were varied by factorial ± irrigation and fertiliser treatments. At Rotorua, P. radiata stem volume after 10 months was greatest in weedfree, gorse, broom, and Yorkshire fog plots and least in herbaceous broadleaf and buddleia plots, with pampas intermediate. At this time, there was no strong evidence of interspecific competition for water or nutrients. At Rangiora, trees growing with grass and herbaceous broadleaves were substantially reduced in stem volume compared to trees in the weed-free, broom, and gorse plots. There was essentially no difference in growth with the latter three treatments.

An intensively pruned and thinned stand of Cupressus macrocarpa Hartweg provided a unique research opportunity for examining aspects of cypress tree growth and utilisation. At 27 years of age, the stand was harvestable in terms of stand volume, log size, and sawlog quality, identifying C macrocarpa as a species with potential for short rotations. A heavy thinning at age 21 years resulted in minimal stem diameter response but caused a large increase in the number of large branches which in turn reduced the unpruned log quality. Based on a current growth model, a simulated Pinus radiata D.Don stand grown under the same regime and site conditions would have produced a similar total stand volume.

Sawing conversion from roundwood for a 39-log sample ranged from 52% in butt logs to 43% in third logs. Butt logs had low recovery of Clears grades owing to large knotty cores, but the inclusion of a weatherboard/exterior finish grade improved the total "high value" grade recovery to 73% of sawn timber. Corresponding recoveries for second and third logs were 13% and 8% respectively, reflecting the presence of large branches and sapwood.

The yield of kraft pulp from top logs and slabwood was poor in comparison to P. radiata, but the chemical consumption was normal. Paper from these pulps had very low bulk owing to collapse of the short low-coarseness fibres and had low tear strength but reasonable light-scattering properties. Cupressus macrocarpa could be used to produce kraft pulp if mixed in a low proportion with other more-favoured species.

Thermomechanical pulp made from slabwood chips was within the accepted range of commercial properties but, compared with P. radiata, energy consumption was high and refiner motor loadings were very unstable. The resulting paper had low brightness and strength. Attempts to make chemi thermomechanical pulp within the accepted normal commercial range of properties were unsuccessful.

In 11.5-year-old stands of E. regnans F. Muell. with highly variable growth, trees of below-average size and health occurred in small discrete clumps suggestive of microsite influence. Microsites were found to have a variable thickness of Taupo tephra overlying earlier Tirau tephra. The Taupo tephra had lower concentrations of phosphorus and of feeding roots than the Tirau tephra. Where Taupo tephra was thicker than 50 cm in the soil profile, basal areas averaged one-third lower than other microsites and correlated positively with phosphorus concentration in the A horizon.

Foliage samples suggested trees on soil with greater than 50 cm of Taupo tephra have more variable concentrations of foliar elements and lower concentration of phosphorus than trees on soil with predominantly Tirau tephra. These differences attributed to the thickness of Taupo tephra accounted for only some of the total growth variation.

In four small isolated stands of Nothofagus solandri var. cliffortioides (Hook, f. Poole (mountain beech), ranging in size from 70 to 140 m² , exposure was the dominant influence on stand architecture. The most exposed stand exhibited progressive canopy collapse which may eventually lead to its extinction. As shelter increased, stands showed a rapid transition to an intact protective canopy. Dramatic differences in environmental conditions were evident between sheltered and exposed stand edges. Seedlings established in the lee of existing stands while collapse occurred at the exposed edge, leading to stand mobility over time.

Regeneration in montane Podocarpus hallii Kirk/Weinmannia racemosa L.f. (Hall's totara/kamahi) forest on Mt Pureora, central North Island, New Zealand, was sampled in the three phases of the forest growth cycle. Average age differed significantly between the gap (12-year), building (67-year), and mature (227-year) phases. Diameter growth rates were similar amongst major canopy species, but significantly faster during the gap and building phases (0.31 cm/year) than the mature phase (0.19 cm/year).

Two broad patterns of successional change were evident between phases, some species populations (e.g., small-leaved Coprosma spp., tree ferns) increasing during the gap phase and declining thereafter, and others (e.g., Hall's totara, kamahi, Griselinia littoralis Raoul (broadleaf), Myrsine salicina Hook.f. (toro)) increasing during the building phase. Pseudowintera colorata (Raoul) Dandy (horopito), although a gap invader, was prominent throughout. Replacement strategies reflect differing shade tolerance, relatively intolerant species being gap invaders and those with intermediate tolerance establishing during the building phase. Because canopy dominants mostly establish later in the succession ("building-phase replacement"), gap composition alone is not a reliable indicator of future canopy composition in this forest type.

Gap-maker/gap-filler and canopy/understorey relationships suggest that a regeneration cycle involving alternation of podocarps and broadleaved trees, similar to that described from nearby lowland forest, may be operating. Although no unequivocal disruption of natural replacement processes in major canopy species is evident, other studies show that progressive impoverishment of lower tiers by introduced browsing mammals is occurring here.

The paper which this Corrigendum refers to is available here: Within- and between-tree variation in microfibril angle in Pinus radiata

The effluent from a medium-density fibreboard mill, using Pinus radiata D.Don as the chief source of raw material, has been analysed for total and suspended solids, particle size distribution, reducing sugars, total carbohydrate, nitrogen, and phosphorus. The waste liquors, which contained many fine cellulose fibres, were moderately biodegradable with BOD₅/COD ratios from 0.49 to 0.22. Microbial utilisation of the raw wastewater was enhanced by supplements of nitrogen whereas addition of phosphate had little effect.

Extraction of Pinus radiata D.Don needle epicuticular wax and processing of the extract to remove the estolides and fatty and resin acids afforded three minute fractions from silica gel column chromatography, representing less than 0.15% of the wax. The fractions were variously shown to contain a mixture of natural products—viz sesquiterpenes, comprising mainly amorphene, cadinenes, germacrene, calamenene, calacorene, and cadalene; a complex mixture of bisnor, nor, and diterpene hydrocarbons; as well as methyl esters of resin acids and 9,10-seco-dehydroabietanes of unknown ring- A structure. The anthropogenic chemicals identified included polycyclic aromatic hydrocarbons, and a variety of organochlorine compounds, hexachlorocyclohexanes lindanes), hexachlorobenzene, chlordanes, p,p'-DDE, and pentachlorobiphenyl, hexachlorobiphenyl, and heptachlorobiphenyls. It is suggested that analysis of P. radiata needle epicuticular wax would provide one means of biomonitoring in order to assess the quality of the New Zealand environment.

Microfibril angle at breast height and at 18 m was assessed for three genetically distinct groups of Pinus radiata D.Don trees. The three groups consisted of five trees of undefined parentage, five trees of open-pollinated NZ850-55, and five trees of NZ850-55 x Guadalupe provenance P. radiata. Microfibril angle within the core wood at breast height was consistently lower for the two NZ850-55 progeny groups than for the corewood of the control trees, while outerwood values were comparable. Angles at 18 m height were similar to those at breast height for the two NZ850-55 progeny groups but were significantly lower in the control trees. These differences have implications for wood and paper properties.

Needle and branch weight were estimated for each 2-m height zone of the crown in Pinus radiata D. Don trees of five size-classes for each year from ages 5 to 13 years. The fraction of needles aged 1 year decreased down the crown and with increasing tree age. The ratio of branch to needle production did not differ significantly with position in crown, tree size-class, or tree age, and averaged 0.75.

Laboratory populations of Cleopus japonicus Wingelmuller (Coleoptera : Curculionidae) were studied in controlled conditions. A constant temperature of about 20°C and a photoperiod of up to 14 hours appeared to optimise survival and oviposition of C. japonicus and indicated that this weevil should readily acclimatise to conditions in New Zealand.

Incidence of Armillaria root disease and populations of the causal fungi Armillaria novae-zelandiae (Stevenson) Herink and A. limonea (Stevenson) Boesewinkel were monitored in four plots (36 x 28-36 m) in a Pinus radiata D. Don plantation on a site converted from an indigenous podocarp-broadleaf forest in the Bay of Plenty district, New Zealand. Armillaria-caused mortality in different plots varied between 22% and 35% of all trees 5-6 years after planting, while total infection ranged from 54% to 64%. Among the final-crop trees only, up to 4% were killed by Armillaria spp. and between 46% and 51 % were chronically infected. In one plot 67% of pine stumps were colonised by Armillaria spp. 15 months after a non-commercial thinning, providing potential supplementary inoculum for further infection among remaining trees. Armillaria isolates from the four plots belonged to 63 vegetative compatibility groups, of which 27 A. novae-zelandiae and 15 A. limonea groups were first identified prior to clearing and burning of the indigenous cover, and 21 were found post-burn only. Most pre-burn groups were redetermined in approximately the same positions after burning, but four were collected from new locations within the same plots. Post-burn groups were all A. novae-zelandiae, and some appear to represent new colonies introduced by means of basidiospores after burning. Potted seedling experiments were used to confirm the pathogenicity to P. radiata of many of ths Armillaria vegetative compatibility groups.

The impact of possums (Trichosurus vulpecula Kerr.) and insects and disease on the canopy development of pohutukawa (Metrosideros excelsa Gaertn.) was studied at Homunga Bay on the Coromandel Peninsula. Possums were the only threat to established trees through damage to foliage and vegetative buds. The study also showed that regeneration occurred rarely because of the presence of feral goats and domestic sheep and cattle. It is recommended that possum control be carried out in late winter so as to protect new vegetative buds, and that fencing out of domestic stock to allow regeneration be done where possible.

Soils under various stockings of Pinus radiata D.Don at the Tikitere Agroforestry Research Area near Rotorua have shown a significant decline in soil pH with increasing tree age and at higher tree stockings. Soil pH levels in 1975 (year 3) for 0,100,200, and 400 stems/ha were 5.6, 5.7, 5.7, and 5.6 respectively, compared with 5.6, 5.4, 5.3, and 5.0 in 1991. Higher phosphorus and sulphate-sulphur levels under trees and an increase in magnesium on open pasture were also evident. A soil profile study at year 19 indicated pH reductions to 150 mm depth, with increasing phosphorus levels to 75 mm depth. Sulphate-sulphur levels increased at all depths in 400 stems/ha plots.

Vector analysis, previously used to study nutrient status in trees, has been modified for use in competition experiments. The interpretation of changes in leaf weight, nutrient concentration, and nutrient content per leaf indicates whether moisture and/or nutrients are causing changes in growth.

An agroforestry trial with Pinus radiata D.Don and six ground cover treatments was studied using the new interpretation. The results indicated that, for this site/season combination, the main competition factors reducing tree growth were moisture, nitrogen, and boron. Potassium and magnesium levels were also changed but not in a manner expected to alter tree growth. There was' little change in phosphorus, calcium, and micronutrients other than boron. The effect of moisture was consistent with the drought conditions experienced during the study. Lucerne proved to be the most competitive of the ground covers used in this trial.

Vector analysis gave more information on effects of competition than did foliar concentration, foliar nutrient content, or tree growth. Fascicle weight was found to be an indicator of tree vigour, but for P. radiata it was not as sensitive as height and diameter measurements.

Internode length is an important variable in determining the amounts of clearwood which can be obtained from unpruned logs. An empirical model has been developed for predicting stand mean internode length for variable log lengths for forest sites in New Zealand. The model, applicable for both unimproved and genetically improved Pinus radiata D. Don, predicts internode length from mean annual rainfall, altitude, and "level of genetic improvement".

For trees which are naturally insect-pollinated and have effective pollination ranges of less than 40 m, the Forward Selection Plot (FSP) design provides an open-pollinated breeding population layout that is effective for ranking families, for within-family selection, and, subsequently, for collection of seed for the breeding population of the next generation. It therefore provides a basic field resource for breeding which will give nearoptimal genetic gain at low cost. It can be used with a sublined breeding population and can also be used to provide improved seed in commercial quantities. The use of the computer programme "NO INCEST" to help layout FSPs minimises the chances of inbreeding in the seed collected after evaluation and roguing of the test.